New artificial intelligence technology can now detect early signs of over 1,000 diseases long before any symptoms appear, according to new research.
The computer algorithm, named MILTON, analyzes patient test results commonly collected by general practitioners to identify patterns in the data. It can predict with high confidence the likelihood of a disease diagnosis many years in advance.
Pharmaceutical giant AstraZeneca, which developed this AI tool, says it will accelerate the creation of more effective and targeted treatments.
Additionally, AstraZeneca is making the data freely available to other researchers, allowing them to develop diagnostic tests that could enable early preventive treatments to halt diseases before they progress.
Lead researcher Slave Petrovski told, “For many of these diseases, by the time they manifest clinically and the individual visits a doctor due to a symptom or visible issue, the disease process may have already been underway for a long time.”
He explained that a cascade of events in the blood may have occurred before symptoms emerge.
“We can identify signatures in an individual that are highly predictive of diseases like Alzheimer’s, chronic obstructive pulmonary disease (COPD), kidney disease, and many others.”
AstraZeneca used data from 500,000 individuals who are part of the UK Biobank, a vast repository of health information.
MILTON analyzed data from 67 routine clinical biomarkers, which included detailed results from blood and urine tests, blood pressure measurements, respiratory function checks, as well as factors such as weight, age, and sex.
Additionally, MILTON examined data from 50,000 Biobank participants on 3,000 proteins found in blood plasma, which play critical roles in many bodily functions, including the immune and hormonal systems.
The AI tool was able to detect subtle patterns in the data that would have been impossible for the human eye to see, and link those patterns to diseases that were diagnosed up to a decade or more later.
The study rated MILTON’s predictive performance as “exceptional” for 121 diseases, and “highly predictive” for another 1,091.
Although routine biomarkers collected by GPs were effective in predicting diseases, adding the analysis of 3,000 proteins—without significantly increasing the cost of blood tests—greatly enhanced MILTON’s ability to flag patients at risk.
Dr. Petrovski highlighted that the discovery opens up the potential for much earlier intervention in the disease process.
“There is always an opportunity to combine lifestyle changes with pharmaceutical interventions to achieve the best health outcomes,” he said.
He emphasized that many diseases are difficult to reverse once they reach later stages, so intervening earlier could help manage and prevent disease progression.
AstraZeneca stressed that MILTON is currently a research tool, and more work is needed before it can be used by doctors in clinical settings.
Professor Tim Frayling, a genetics expert at the University of Exeter, was cautious about the findings.
He said, “We need to be careful when claiming we can ‘predict disease,’ when we actually mean ‘we can give you a slightly better understanding of your chances of developing a disease, but many unknown factors remain.’
This approach will likely improve our knowledge of disease development rather than definitively predict who will develop them.”
Professor Dusko Ilic of Stem Cell Sciences at King’s College London praised MILTON’s predictive power as “remarkable,” suggesting it opens the door to earlier interventions, more personalized treatments, and reduced healthcare costs.
However, he voiced concerns about the ethical use of such a powerful tool. “Unregulated, MILTON could be misused by health insurance companies or employers to assess individuals without their consent, leading to discrimination and breaches of privacy.”
Ilic stressed the importance of strict guidelines and oversight to ensure MILTON’s benefits are realized ethically and responsibly.
